FoodButler

Inspiration

We always had problems choosing what to eat and wished there was a way to remove to necessity to choose. So we came up with a bot that would give us food recommendations through NLP that will enable us to be able to remove the burden of choice.

What it does

Users login to talk to a bot, and we used Named-entity recognition to detect user's food choices and recommend them places based on their location and how likely it falls within their interests.

How we built it

We used React for the frontend, SocketIO for the chatbot messaging, Flask for our API routes,

Challenges we ran into

It was challenging to try out the NER although we were not able to successfully come up with the keyword recognitions, it was a good experience.

Accomplishments that we're proud of

We are proud that we were close to completing the project given that it is our first time working together and using new technologies on the fly.

What we learned

We learnt about the issues of SocketIO during deployment although everything worked perfectly when running on localhost. We managed to fix the error due to Procfile configuration.